bigdatamesosdcosjupyter-lab

Allocate a mesos service using more than 1 node resources

I have a DC/OS mesos cluster with 1 master, 2 private agents and 1 public agent. Each private agent offers 4 CPU and 14.7 GB memory. The problem is: How could I allocate a service that needs 8 CPUs and 20 GB? Mesos tries to allocate the service in one node, so I would like to join the slaves resources to run this service. (Jupyterlab for more info)

Thanks!

Solution

  • That’s simply not possible. You cannot create services that use more resources than at least one single agent can offer.

    Please check e.g. http://datastrophic.io/resource-allocation-in-mesos-dominant-resource-fairness-explained/ to get an idea on how resource allocation works with Mesos.